Streaming camera using flask API while being able the start and stop recording of the video

[KarasuY]

I've managed to get this working successfully with picamera but I've been trying for weeks to get it working on picamera2.
I've gotten to the point where I can start a stream that records simultaneously but I can't separate the 2. I need to be able to open and close viewing the stream while having it continue to record until I send a command to stop recording. Here is the code that records and streams simultaneously, I'd really appreciate any help fixing this issue. I've already looked through every bit of relevant documentation including the picamera2 manual and specifically section 9.3 that should technically be the solution.

from ffmpegoutput_mp4 import FfmpegOutput #customized version that ensures that the output is mp4 instead of mp4v

import picamera2 #camera module for RPi camera
from picamera2 import Picamera2
from picamera2.encoders import JpegEncoder, H264Encoder
from picamera2.outputs import FileOutput #, FfmpegOutput
import io

import subprocess
from flask import Flask, Response
from flask_restful import Resource, Api, reqparse, abort
import atexit
from datetime import datetime
import threading
from threading import Condition
import time 
from random import randint
import boto3
import imageio 

S3_BUTCKET = 'dct-cambucket1'
s3 = boto3.client('s3')

app = Flask(__name__)
api = Api(app)

class Camera:
	def __init__(self):
		print('hi')
		
		self.fileOut = None
		self.output_filename = ''
	def get_frame(self, start, ID):	
		if start:
			now  = datetime.now()
			timestamp = now.strftime("%Y-%m-%d_%H-%M-%S")
			self.output_filename = f'{ID}_{timestamp}.mp4'
			self.camera = picamera2.Picamera2()
			self.camera.configure(self.camera.create_video_configuration(main={"size": (640, 480)}))
			self.encoder = JpegEncoder()
			self.fileOut = FfmpegOutput(self.output_filename, audio=False) #StreamingOutput()
			self.streamOut = StreamingOutput()
			self.streamOut2 = FileOutput(self.streamOut)
			self.encoder.output = [self.fileOut, self.streamOut2]
			
			self.camera.start_encoder(self.encoder) 
			self.camera.start() 
			
		
		with self.streamOut.condition:
			self.streamOut.condition.wait()
			self.frame = self.streamOut.frame
		
		return self.frame
		
				
	def stop_recording(self):
		self.fileOut.stop()
		self.camera.stop()
		self.camera.close()
		print('recording stopped')
		return True
	
	def upload_file(self):
		print(self.output_filename)
		subprocess.run(['aws', 's3', 'cp', self.output_filename, f's3://dct-cambucket1/{self.output_filename}'])
		print('Uploding complete')
		return self.output_filename
		
class StreamingOutput(io.BufferedIOBase):
	def __init__(self):
		self.frame = None
		self.condition = Condition()

	def write(self, buf):
		with self.condition:
			self.frame = buf
			self.condition.notify_all()
		
#defines the function that generates our frames
camera = Camera()

def genFrames(ID):
	frame = camera.get_frame(True, ID)
	while True:
		frame = camera.get_frame(False, None)
		yield (b'--frame\r\n'
			   b'Content-Type: image/jpeg\r\n\r\n' + frame + b'\r\n\r\n')

            
#defines the route that will access the video feed and call the feed function
class VideoFeed(Resource):
	def get(self, ID):
		print('Recording Started')
		return Response(genFrames(ID),
                    mimetype='multipart/x-mixed-replace; boundary=frame')

# class StartRec(Resource):
	# def get(self):
		# camera.start_recording()
		# return {'Status': 'True'}
		
class StopRec(Resource):
	def get(self, ID):
		camera.stop_recording()
		filename = camera.upload_file()
		link = "https://dct-cambucket1.s3.amazonaws.com/" + filename

		return {'status': 'True', 'message': 'Recording Stopped', 'ID': link}, 201
		
		
api.add_resource(VideoFeed, '/picam0002/cam/<ID>')
# api.add_resource(StartRec, '/picam/start_rec/<ID>')
api.add_resource(StopRec, '/picam0002/stop_rec/<ID>')

if __name__ == '__main__':
	app.run(debug = False, host = '0.0.0.0', port=5000)

[IcyG1045]

I've been working on it here and have pretty much solved your issue. I hope that it'll work on your setup because I'm using a CM4 but I don't see why it shouldn't work on another pi running bullseye 32bit.

I am currently working on getting file thumbnails generate and to get auto start/stop recording from motion detection.

I've also got on the fly .h264 to .mp4 conversion working as well as a live file viewer. The file conversion is not on current code but I will update it when I get home from work.

https://github.com/IcyG1045/CM4Cam